Vincent Porretta

Vincent Porretta

Data Scientist

Amazon

As a data scientist at Amazon Alexa, I work in the Generative AI (GenAI) and Large Language Model (LLM) space, focusing on improving human-AI interactions at scale.   

Interests

  • Psycholinguistics
  • Spoken language processing
  • Foreign-accented speech
  • Visual World Paradigm eye-tracking
  • Second language acquisition and Bilingualism

Education

  • PhD in Linguistics, 2015

    University of Alberta

  • MEd in Foreign and Second Language Teaching, 2007

    Ohio State University

  • MA in Italian, 2006

    Ohio State University

  • BA in Italian, 2004

    Ohio State University

Skills

Coding

Python, Streamlit, R, Shiny, R Markdown

Agentic AI

Prompting, Train/Test, Optimization

Statistics and Data Science

LMER, GAMM, Time-series Analysis, Clustering

Languages

English, Italian, Spanish, Portuguese, Finnish

Linguistic Research

Language processing, Acoustic analysis

Teaching and Learning

Pedagogy, Curriculum Development

Experience

 
 
 
 
 

Data Scientist

Amazon

Feb 2025 – Present Boston, MA
 
 
 
 
 

Language engineer

Amazon

Sep 2020 – Jan 2025 Boston, MA
 
 
 
 
 

Sessional faculty

McMaster University

Sep 2019 – Apr 2020 Ontario
 
 
 
 
 

Sessional faculty

Brock University

Sep 2019 – Dec 2019 Ontario
 
 
 
 
 

Postdoctoral research fellow

University of Windsor

Jun 2017 – May 2019 Ontario
 
 
 
 
 

Postdoctoral research fellow

University of Alberta

Sep 2015 – May 2017 Alberta
 
 
 
 
 

Foreign Language Teacher

Bishop Hartley HS

Sep 2007 – Aug 2010 Ohio

R packages

PupilPre

PupilPre: Preprocessing Pupil Size Data

VWPre

VWPre: Tools for Preprocessing Visual World Data

Teaching

University Teaching

University-level courses taught

Workshops

Invited workshops taught

Collaborators

Lines between names indicate coauthorship on publications. The darkness of the lines indicates the number of joint publications.
Collaborators

Publications

Quickly discover relevant content by filtering publications.

Speaker-specific cues influence semantic disambiguation

Addressees use information from specific speakers’ previous discourse to make predictions about incoming linguistic material and to …
Catherine Davies, Vincent Porretta, Kremena Koleva, Ekaterini Klepousniotou
DOI

A lingering question addressed - Reading rate and most efficient listening rate are highly similar

Is it possible that silent reading rate is the same as the most efficient listening rate? The hypothesis has been formulated in the …
Victor Kuperman, Aki-Juhani Kyröläinen, Vincent Porretta, Marc Brysbaert, Sophia Yang
DOI

(Mis)understanding your native language - Regional accent impedes processing of information status

Native-speaker listeners constantly predict upcoming units of speech as part of language processing, using various cues. However, this …
Anja Arnhold, Vincent Porretta, Aoju Chen, Saskia A. J. M. Verstegen, Ivy Mok, Juhani Järvikivi
DOI
See all publications

Presentations

When 'bug' creeps into memory - False memories and foreign-accented speech

2019-11-14 13:00 — 13:00 Montreal, QC, Canada
Vincent Porretta, Aki-Juhani Kyröläinen, Ghadir Nassereddine, Lori Buchanan

Punny and funny - Semantic association in pun reading

2019-06-07 13:00 — 13:00 Waterloo, ON, Canada
Taylor Gooding, Vincent Porretta, Lori Buchanan

Investigating the comprehension and perception of reduced speech with pupillary response

2019-05-13 13:00 — 13:00 Louisville, KY, USA
Benjamin v. Tucker, Vincent Porretta, Yoichi Mukai
See all talks

Topics

Accent Accented Speech Erp Eye-Tracking Fixation Durations Gamm Information Status Invited Language Processing Lexical Competition Listener Experience Non-Refereed Perception Prediction Pupil Dilation Rating Reaction Times Second Language Learning Semantics Vwp

Contact

Feel free to message me via LinkedIn

  • Amazon, 55 Pier 4 Boulevard, Boston, MA 02210